~ruther/stm32h747i-disco-usb-image-viewer

c298348f — Rutherther 10 months ago
few minorities
6107f1b1 — Rutherther 10 months ago
chore: split main functions
f31e7600 — Rutherther 10 months ago
fix: ability to change pll vcosel, disable pll outputs
8313b61d — Rutherther 10 months ago
feat: implement fmc sdram control
b41c1c46 — Rutherther 10 months ago
feat: implement clocks enabling
4c34bf99 — Rutherther 10 months ago
fix: unicode strings
76bfb768 — Rutherther 10 months ago
Revert "feat: atomic register modifications"

This reverts commit 997dd77e0476ee3929fdab6b1ad29cef85f32c2b.
079ff751 — Rutherther 10 months ago
feat: implement cdc application layer, fix enumeration
28319543 — Rutherther 10 months ago
feat: atomic register modifications
d5bfe1a3 — Rutherther 10 months ago
fix: usb should not rewrite DIEPCTL bits
60e86412 — Rutherther 10 months ago
feat: add rx data callback, increase tx fifo size
f4b4b64a — Rutherther 10 months ago
fix: make queue thread safe
cb4fa519 — Rutherther 10 months ago
fix: ensure register access are inline
a42a0f99 — Rutherther 10 months ago
tests: add device descriptor test

This is just an initial skeleton of
how the tests could look like.
It's proving hard to think of a good
way to encapsulate the tests. Every
test will need a different set of files
from the project compiled, as some
files should be stubs instead. Additionally,
those stubs could have some commonalities.
It should for example be possible to initialize
the usb device from the tests. It would be good
to implement at least some basic handling
of the usb registers as well. - ie. if you
enable an endpoint and put in data through fifo,
the endpoint should be disabled. It should be
possible to send data from the host to the device
as well - reading from fifo. And stuff like that...
0207e427 — Rutherther 10 months ago
chore: fix packing on x86_64
8776612f — Rutherther 10 months ago
chore: add gcc toolchain to compile tests
89d71e0e — Rutherther 10 months ago
fix: correctly send functional descriptor

The functional descriptor was not sent correctly,
specifically, it was pointed to its address,
instead of the pointer itself. That meant bunch
of irrelevant stuff was sent instead.
9db48e96 — Rutherther 10 months ago
feat: use task_result to indicate blocking

There should be no spin loops inside of the
interrupts, so this change removes them,
and instead uses a result system that has
WOULD_BLOCK result. When this result is returned,
it is expected there was no side-effect and the
function can safely be reran later.
5d5e3d6c — Rutherther 11 months ago
feat: implement enumeration fully
f8fd8e9f — Rutherther 11 months ago
chore: update pin, exti, delay
Next